Placement Algorithm resettlement

Using machine learning to predict where resettled refugees are likely to thrive

Airbel is piloting and scaling an innovative machine learning algorithm developed by the Stanford University Immigration Policy Lab that matches refugees in areas where they are likely to thrive when resettled.
Filter by status
Design
Test
Scale

Placement Algorithm is at the Generate Solutions stage

financing

Displacement Insurance

Displacement insurance disburses funds for swift life-saving action during crises, with an overall aim to make humanitarian aid more agile.

Find out more

resettlement

Resettlement Finance

In this new model for refugee resettlement finance, a self-sustaining fund provides countries with finance for refugee resettlement, and is replenished as refugees fuel economic growth.

Find out more

resettlement

Placement Algorithm

U.S.

Airbel is piloting and scaling an innovative machine learning algorithm developed by the Stanford University Immigration Policy Lab that matches refugees in areas where they are likely to thrive when...

Find out more

education

Pop Up Learning

Bangladesh

Pop-Up Learning is a computer-assisted learning program, aimed to mobilize quickly and efficiently on the onset of crisis.

Find out more

employment

Trade Accelerator

Focused support for Jordanian businesses to foster rapid growth and help them tap into the European market and hire refugees to do so.

Find out more

This project became inactive following the Generate Solutions phase

nutrition

Virtual Baby

This portable game would let children earn points for raising a happy baby, teaching them about sexual health and nutrition.

Find out more

This project became inactive following the Generate Solutions phase

resettlement

Ask Laila

A hotline, SMS, and web-based service to help resettled refugees get quick answers from a trusted local source - in their own language.

Find out more

This project became inactive following the Prototype phase

nutrition

Sprout

Niger

Over 16 million children under the age of 3 are afflicted by severe acute malnutrition. To pinpoint the most impactful prevention methods, Sprout is experimenting with 3 distinct approaches.

Find out more

education

Spark

Tanzania

Spark creates alternative learning spaces outside of refugee camp schools to practice play-based learning and social-emotional skills with primary school children.

Find out more

This project became inactive following the Prototype phase

resettlement

Caseworker Communications

United States

In partnership with Twilio.org, Airbel is developing an SMS-based system to streamline communication between caseworkers and newly-settled refugees as they adapt to their new lives in the United States.

Find out more

education

Coach Everu

Tanzania

Coach Erevu is a digital coaching program that aims to build the social emotional skills of teachers through weekly peer-to-peer video clubs, timely habit triggers, and bite-sized teaching materials

Find out more

nutrition

Low-Literacy Malnutrition Tools

South Sudan

In hard-to-reach areas local community health workers are key to treating malnourished children. Commonly used tools for diagnosing and treating malnutrition have been adapted to meet the needs of workers...

Find out more

This project graduated following the Pilot phase

employment

Match

Jordan

Working to counter the high rates of unemployment that follow mass displacement, the Match project in Jordan is testing a set of new innovations for placing vulnerable populations and refugees...

Find out more

education

Sesame Seeds

In this historic effort, IRC and Sesame Workshop help children whose lives have been uprooted by the Syrian war to build brighter futures.

Find out more

This project graduated following the Pilot phase

education

Tunakujenga

Tanzania

Tunakujenga is a family learning program that empowers caregivers to learn and practice social and emotional learning activities with their children at home.

Find out more

safety

Becoming One

Uganda

Becoming One is a counseling program designed to bring couples closer and prevent intimate partner violence. The program is delivered in partnership with local faith leaders, who guide couples through...

Find out more

safety

Modern Man

Liberia

Modern Man is a mobile messaging campaign that engages men with an aspirational and positive masculine identity to prevent intimate partner violence. It sends SMS messages to men who have...

Find out more
There are currently no projects in the Scale phase

Moving somewhere new is challenging under the best circumstances. It’s especially difficult for refugees, who are driven from their homes by conflict or major threats. Refugee resettlement offers one of the most transformative opportunities to those affected by conflict. When refugees are resettled in the United States, resettlement agencies like the International Rescue Committee determine where to send refugees and caseworkers have played a critical role in helping refugees adapt to new places.

The community where refugees are placed is generally dictated by need and availability. But data shows what common sense states: different people are likely to succeed in different circumstances. Enter the placement algorithm - a collaboration between Stanford and the Airbel Center that analyzes historical data on refugee demographics, local market conditions, individual preferences and outcomes to generate predictions that suggest an ideal location for resettled refugees. This actionable information can then be used to inform decisions about where to place refugees in the U.S. We ultimately aim to scale this approach to help identify where we can place refugees globally.  

Placement Algorithm is part of a larger enterprise to revolutionize refugee resettlement, by harnessing private capital, data and volunteers to change the calculus for host countries in determining whether they resettle—and enable many more refugees to start a new life in a welcoming country. Other components include Resettlement Finance, a self-sustaining fund provides countries with finance for refugee resettlement, as well as a digital co-sponsorship platform, that mobilizes sponsors to welcome refugees as they arrive. Together, these various components aim to positively change the experience for refugees and host countries. Ultimately, this more sustainable model could increase the volume of refugees resettled globally.­

June 2018 | Cross-agency partnership towards a pilot

The IRC will work with Lutheran Immigration and Refugee Service and other potential resettlement agencies to rollout a pilot of the algorithm.

Resource

March 2018 | “Moonshot” pitch to a16z

Matching becomes one of the cornerstones of how we pitch a “global resettlement” vision; now, we want to match both within countries, and across them.

Resource

January 2018 | Stanford publishes algorithm results

Stanford completes a backtest of the algorithm on IRC data, and the results are incredibly positive, with certain simulations leading to a doubling of employment at 90 days for resettled refugees.

Resource

November 2017 | Matching algorithm emerges among top ideas

After design workshops, brainstorms, and a look at IRC’s US Resettlement outcomes data, IRC and Stanford’s IPL convene to prioritize among the early ideas. The Matching Algorithm is treated as a “no-brainer.”

Resource

August 2017 | Design sprints in US Resettlement offices

Design sprints take place in Oakland and San Diego offices, interviewing refugees, caseworkers, and volunteers.

June 2017 | IRC / Stanford Collaboration kicks off

This represents one of the first big Airbel partnerships with an academic institution and design thinking capabilities.

Resource

We develop new solutions to better serve people suffering the effects of conflict and crisis.